A novel gene expression signature is able to distinguish between metastasizing and nonmetastasizing cutaneous squamous cell carcinoma (cSCC) in patients identified as having low-risk disease by standard clinicopathological staging systems.

The new signature identified approximately one sixth of patients with T1 disease as having a metastatic risk comparable to that of patients with T2a disease. The patients with T1 disease identified as having a metastatic risk comparable to that of patients with T2a disease accounted for approximately 70% of metastatic events in the T1 population. Among patients with T2a disease, “a high-risk subset of about 1%” accounted for approximately a quarter of the metastases in this group, according to the study abstract.

This is the “largest and most comprehensively annotated squamous cell carcinoma for RNA sequencing” to date, said Marlies Wakkee, MD, PhD, while presenting the findings of the Dutch analysis at the European Association of Dermato-Oncology (EADO) Congress 2026. 

Wakkee explained that the currently available profiles are primarily based on low numbers of patients, typically less than 100, and often are derived from academic, expert centers, “where they include mainly high-risk patients [based on clinicopathological staging].”

All the patients in the new study are ‘low risk’ in terms of clinical staging, but metastatic disease develops in 2%-5% of patients with cSCC, said Wakkee, who is a dermato-oncologist at Erasmus MC Cancer Institute, University Medical Center in Rotterdam, Netherlands. This new gene signature score is an attempt to find that particular group of metastatic patients, because the current staging systems fail to identify all patients at high risk for metastasis, she continued.

Wakkee underlined that, on one hand, there is a need to identify those patients who are truly at low risk for metastatic disease on their gene signature so they can have less follow-up, and, on the other hand, to identify patients at higher risk for metastasis “so we can offer them more personalized treatments.”

Methodology and Results

To develop the new biomarker, Wakkee and colleagues conducted the D-SQUAME study, which included a nationwide discovery cohort of 19,120 patients. From this, 183 cases with metastatic disease and the same number of nonmetastasized control cases were selected to perform a nested case-control analysis.

Primary tumor samples underwent bulk RNA sequencing, with 60,295 genes initially identified and filtered down to 11,144 candidate genes, including those already known from the literature alongside a data-driven approach to pick out novel genes.

Modeling analysis narrowed that down to 23 genes covering differentiation, stemness, homeostasis, and proliferation in a regularized Cox regression model, which yielded a novel gene expression signature termed SSCore-GEP.

The new gene signature was then validated in a separate set of 51 cases and 51 control individuals selected from a nationwide cohort of 25,921 cSCC patients, and in 22 cases and 30 control individuals from a recently published study by Nassir and colleagues.

In the D-SQUAME validation cohort, the SSCore-GEP outperformed standard clinicopathological staging systems, at a weighted C-index of 0.83 vs 0.66 for the Brigham and Women’s Hospital (BWH) staging system and 0.63 for the American Joint Committee on Cancer 8th edition.

In Nassir and colleagues’ validation cohort, SSCore-GEP had an area under the receiver operating characteristics curve of 0.79 vs just 0.45 for the BWH staging system.

“Of course, more studies are needed to bring it to the clinic,” Wakkee noted, particularly to include more immunocompromised patients, as well as to have larger patient populations and those with higher risk disease.

Analytical validation will also be needed to determine “a really helpful risk threshold to change your treatment plan, for example,” Wakkee continued. “And of course, cost-effectiveness studies are needed.”

Q&A on Findings

Following the presentation, Catherine Harwood, PhD, who co-chaired the session, asked Wakkee how she saw the novel signature being used as part of the clinical workflow for cSCC.

It could potentially be directed to high-risk patients in terms of clinical stage, or in certain subgroups, such as immunocompromised patients or those with particular disease subtypes, Wakkee said.

The study presenter pointed out that, currently, treatment decisions are based on biopsy results, “and we all know that biopsy is usually not representative,” but simply “tells you it’s a squamous cell carcinoma.”

With many disease characteristics changing during the course of treatment, Wakkee speculated that SSCore-GEP could be used to identify, for example, candidates for neoadjuvant therapy.

Harwood, Clinical Professor of Dermato-Oncology, King’s College London, London, England, also asked whether there are sufficient numbers of immunocompromised patients in their cohort to examine the impact of the novel signature.

Wakkee said that it does contain patients who have undergone an organ transplant, and they were able to link their data to the national organ transplant registry, and those with hematologic malignancies. However, she conceded that, when looking into subgroups, “you’re easily underpowered” for drawing a firm conclusion.

The project was co-funded by the PPP Allowance made available by Health Holland, Top Sector Life Sciences & Health, an unrestricted research grant from Stichting Hanarth Fonds, an independent donation by Alphatron, and the Erasmus MC Foundation. No relevant financial relationships were declared.